How much twist on a propeller?

[seandepagnier]

TL;DR Summary

how to calculate it?

I am 3d printing micro wind turbine blades (5-10 watts output in 30mph wind), but anyway I am using naca foil shape twisted to make it.

This is not technically a propeller (cannot be efficiently used to push air) because it is generating power from moving air: it is a repeller, but regardless the theory is the same for either, and I need to understand this for making propellers as well.

My question is, how much twist should it have, and how would I calculate the twist correctly?

I have now a linear twist, so starting at 40 degrees at the hub, out to 85 degrees at the tip and it spins really fast in the wind, but it can't be optimal.

When I really think about it, it seems that the twist along the length is non-linear. Meaning it should twist more and more closer to the hub with less rate of change of twist toward the tip. Is this correct? Is it 1/x ? If so, what is the starting twist angle supposed to be?

Further, should the camber or foil shape itself also vary along the length? It seems it should, but how do I calculate this?

 

[jrmichler]

The quick, simple, and inferior, method is to use an airfoil cross section and design for constant angle of attack at your design condition. The much better approach is to search optimal wind turbine blade design. That search found a number of good hits, one of which is: https://www.mdpi.com/1996-1073/5/9/3425/pdf. That PDF is highly recommended reading, even if you choose to use the quick and simple method.

 

[seandepagnier]

I was never using constant angle because this is obviously not going to work well, but I was using a constant twist by twisting linearly which worked pretty well but is not optimal.

Now I have done it piecewise and use the twist angle of atan(1/(tsr*x)) where x is from 0 at hub to 1 at tip and tsr is the tip speed ratio. This gives much more rate of change of twist near the hub, but it's also interesting to see the shape change with different tip speed ratios. I am using 24 segments and it looks smooth. It will be interesting to try it.

I think I can do maybe even better and I"m also wondering about the blade flexing under load, I also need to round the tips, so there is still a lot to experiment